AUTOMATED DISCOVERY OF TELIC RELATIONS FOR WORDNETBackground TermsBackground Terms (cont)Goal of ArticleGeneral algorithmThe Algorithm (Pt I: finding the telic word)PowerPoint PresentationSlide 8The Algorithm (pt II: finding the correct synset)Slide 10Slide 11Slide 12Slide 13Slide 14The Algorithm in ReviewFinal ResultsFuture work.AUTOMATED DISCOVERY OF TELIC RELATIONS FOR WORDNETMARCO DE BONISURESH MANANDHARBackground TermsGlosses – Every word and synset in WordNet contains a short (about one sentence long) description called glosses.Telic Relations - The goal / function of an object. The purpose an agent has in performing an act or the built-in function.IntroductionBackground Terms (cont)Telic Relations (cont) – Some examples of telic relations,•The telic of “milk” might be “drink”.•The telic of “wood” might be either “burn” (make a fire) or build (make furniture). - Objects may have none, just one, or many telic relationships. IntroductionGoal of Article Telic relations are talked about in WordNet (5 papers, pg. 18, 22 ), but never actually implemented. The Goal of the article is to present an algorithm to automatically discover telic relations of words and synsets by looking at glosses (the descriptions of words).IntroductionGeneral algorithmFor every noun in WordNet1. Parse out a telic relation(s) from the glosses for the noun.2. Since the telic word(s) will usually have many senses, find the appropriate synset to match up the noun to.IntroductionThe Algorithm (Pt I: finding the telic The Algorithm (Pt I: finding the telic word)word)- By looking at special patterns (specific - By looking at special patterns (specific wording) within the glosses, telic words wording) within the glosses, telic words can be parsed out. can be parsed out.- The patterns used were - The patterns used were PatternPatternEx. WordEx. WordGlossGlosstelictelic“… “… to TELIC_VERB to TELIC_VERB by use of …”by use of …”MammographMammographyy““a diagnostic procedure to detect a diagnostic procedure to detect breast tumors breast tumors by use ofby use of X rays.” X rays.”Detect breast Detect breast tumorstumors“… “… used for used for Telic.”Telic.”Tracing paperTracing paper““a semitransparent paper used for a semitransparent paper used for tracing drawings.”tracing drawings.”Trace drawingsTrace drawings“… “… used to Telic.”used to Telic.”Cardiac Cardiac glycosideglycoside““Obtained from a number of plants Obtained from a number of plants and used to stimulate the heart …”and used to stimulate the heart …”Stimulation of the Stimulation of the HeartHeart“… “… use of … to use of … to Telic.”Telic.”TrickeryTrickery““the use of tricks to deceive the use of tricks to deceive someone.”someone.”To deceive To deceive someonesomeonePatternPatternEx. Ex. WordWordGlossGlosstelictelic“… “… used as … in used as … in TELIC_ING-VERB.”TELIC_ING-VERB.”PlasticinePlasticine“… “… resembling clay; used as a resembling clay; used as a substitute for clay or wax in substitute for clay or wax in modeling.”modeling.”ModelingModeling“… “… used in used in TELIC_ING-VERB.”TELIC_ING-VERB.”Seal oilSeal oil“… “… from seal blubber; used in making from seal blubber; used in making soap and dressing leather and as a soap and dressing leather and as a lubricant.”lubricant.”Making soap, Making soap, dressing leather dressing leather and lubrification.and lubrification.“… “… used in … as a used in … as a TELIC.”TELIC.”Giant taroGiant taro““Large evergreen … used in wet Large evergreen … used in wet warm regions as a stately warm regions as a stately ornamental.”ornamental.”Use as a stately Use as a stately ornamental.ornamental.“… “… for use as for use as TELIC.”TELIC.”HouseboatHouseboat““a barge that is designed and a barge that is designed and equipped for use as a dwelling.”equipped for use as a dwelling.”Dwelling.Dwelling.“… “… for use in … for use in … TELIC_ING-VERB.”TELIC_ING-VERB.”WherryWherry““light rowboat for use in racing or for light rowboat for use in racing or for transporting goods and passengers in transporting goods and passengers in inland waters and harbors.”inland waters and harbors.”Racing, Racing, transportation of transportation of passengers.passengers.- The patterns used were - The patterns used were The Algorithm (Pt I: finding the telic The Algorithm (Pt I: finding the telic word)word)Problems.Problems. - Sometimes its hard to pick one - Sometimes its hard to pick one specific noun phrase or verb phrase specific noun phrase or verb phrase for the telic word.for the telic word. - Over-generalization: When words - Over-generalization: When words like “be”, “do”, “make”, “thing” were like “be”, “do”, “make”, “thing” were the telic relations, more specific the telic relations, more specific words had to be found.words had to be found.The Algorithm (Pt I: finding the telic The Algorithm (Pt I: finding the telic word)word)The telic word(s) of an object will most likely belong to more than one synset. The second part of the algorithm is to determine the appropriate synset for the object by looking at the distance (or difference) between different concepts. There are many different approaches that can be use to find distance between concepts ( for example wnconnect). The authors measured the relationship between two concepts with “Sematic distance.”The Algorithm (pt II: finding the correct synset)Given a object w and its telic t, the semantic distance takes all glosses from each sense of t and compares it to the gloss of w (the object). The semantic distance is a sigmoid function that applies an number to each sense of t (the telic). To calculate the actual sematic distance the authors look at all the words surrounding the glosses in t (the telic), and words in gloss w (the object).The Algorithm (pt II: finding the correct synset)Semantic DistanceGiven object w with gloss GWw, it has a telic t with T representing all possible synsets (senses) of t. We want to find the correct synset, ts T, by: ts = Argmaxts’T sd( GWw, ts’)* Where sd( a, s ) calculates the semantic distance between sentence a and word w.The Algorithm (pt II: finding the correct synset)Finding the Correct Synset ts*I